19.5 Charcoal Analysis in Sediments
and weaknesses of each method. Agreement in vegetation reconstructions between methods appeared
to decrease with distance between pollen and midden sites. In addition, disparity in results between
the two methods may arise from the restriction of
midden data to rocky habitats and its relatively
sparse geographic coverage.
3. Anderson and Van Devender (1995) examined fossil pollen from several middens and compared the results to an existing macrofossil analysis in northwest Sonora, Mexico, an area with few
suitable sources of pollen from sediments. Pollen
analysis from middens added both local and regional taxa to the suite of pollen types identified
from macrofossils.
19.5 Charcoal Analysis
in Sediments
Changes in the abundance of microscopic and
macroscopic charcoal particles in lake and wetland
sediments can be analyzed to determine fire frequency (Patterson et al., 1987). The temporal resolution of this method, like pollen analysis, depends on the sediment interval sampled and ranges
from annual records in varved lakes to intervals of
hundreds of years for very long sediment records.
The spatial extent of fires inferred from charcoal
analysis is a function of lake or wetland size, in
which small lakes collect charcoal from a local area
while large lakes receive charcoal from a region
(Whitlock and Millspaugh, 1996). In addition,
larger charcoal particles are assumed to be transported shorter distances in the atmosphere than
smaller particles and thus indicate occurrence of
fire in a more local area (Whitlock and Millspaugh,
1996). Quantification of charcoal particle abundance is frequently conducted in conjunction with
pollen analysis from sediment cores, with chronologies established by varve counting or radiocarbon or lead 210 dating (Whitlock and Millspaugh,
1996). Intervals of abundant charcoal are interpreted as evidence of fire events (Millspaugh and
Whitlock, 1995). Examination of pollen records in
conjunction with charcoal analyses provides information about vegetation response to fire.
Reconstruction of fire histories using charcoal
analysis is often combined with dendroecological
methods, in which comparisons of data from increment cores and cut sections of nearby fire-scarred
trees are used to date local fires (Clark, 1990). Both
methods have advantages and limitations. Fire-scar
279
chronologies are spatially more precise than charcoal analysis, but records become more sparse with
time and do not extend beyond tree life-spans, although records from snags and logs can lengthen
chronologies (Clark, 1990). Charcoal analyses can
provide temporal sequences of thousands of years,
although generally with less precise resolution than
fire-scar chronologies (Fall, 1997); they can also be
directly correlated with pollen records (Whitlock and
Millspaugh, 1996). Comparison of fire-scar dates
with charcoal profiles can be used to calibrate thresholds below which charcoal amounts are considered
background levels (Millspaugh and Whitlock, 1995).
Charcoal peaks above the threshold are interpreted
as fire events. Charcoal analysis can also provide an
important source of information about fire frequencies in systems in which trees are killed by fire and
thus do not provide fire scars for dating fire events
(Fall, 1997).
A number of studies have demonstrated a close
relationship between peaks in various charcoal particle size classes and occurrence of local fires (Whitlock and Millspaugh, 1996). However, no measure
of charcoal abundance has provided complete accuracy in detecting known local fires. No charcoal size
examined by MacDonald et al. (1991) consistently
identified all known local fires at Wood Buffalo National Park, Alberta, although the abundance of microscopic charcoal was found to be influenced by
regional fires. Charcoal deposition patterns are a
function of a number of factors, including the size,
intensity, fuel type, and meteorological conditions of
fires, as well as lake and watershed characteristics
(Patterson et al., 1987; MacDonald et al., 1991;
Whitlock and Millspaugh, 1996). In addition, charcoal studies have employed a variety of analytical
techniques. Given this variability in techniques, deposition patterns, and results, Whitlock and
Millspaugh (1996) recommend that different charcoal size classes and analytical techniques be compared in a fire history study to determine the best
method for detecting local fires in an area.
EXAMPLES
1. Millspaugh and Whitlock (1995) constructed
a 750-year fire history for the Central Plateau of
Yellowstone National Park. Regional and local
fires were distinguished by comparing the charcoal
record of a large lake with those of nearby small
lakes. In addition, local fires (within the watershed)
were discriminated from regional fires in charcoal
records by determining magnetic susceptibility,
which is high in burned soils eroded into lake sed-
and weaknesses of each method. Agreement in vegetation reconstructions between methods appeared
to decrease with distance between pollen and midden sites. In addition, disparity in results between
the two methods may arise from the restriction of
midden data to rocky habitats and its relatively
sparse geographic coverage.
3. Anderson and Van Devender (1995) examined fossil pollen from several middens and compared the results to an existing macrofossil analysis in northwest Sonora, Mexico, an area with few
suitable sources of pollen from sediments. Pollen
analysis from middens added both local and regional taxa to the suite of pollen types identified
from macrofossils.
19.5 Charcoal Analysis
in Sediments
Changes in the abundance of microscopic and
macroscopic charcoal particles in lake and wetland
sediments can be analyzed to determine fire frequency (Patterson et al., 1987). The temporal resolution of this method, like pollen analysis, depends on the sediment interval sampled and ranges
from annual records in varved lakes to intervals of
hundreds of years for very long sediment records.
The spatial extent of fires inferred from charcoal
analysis is a function of lake or wetland size, in
which small lakes collect charcoal from a local area
while large lakes receive charcoal from a region
(Whitlock and Millspaugh, 1996). In addition,
larger charcoal particles are assumed to be transported shorter distances in the atmosphere than
smaller particles and thus indicate occurrence of
fire in a more local area (Whitlock and Millspaugh,
1996). Quantification of charcoal particle abundance is frequently conducted in conjunction with
pollen analysis from sediment cores, with chronologies established by varve counting or radiocarbon or lead 210 dating (Whitlock and Millspaugh,
1996). Intervals of abundant charcoal are interpreted as evidence of fire events (Millspaugh and
Whitlock, 1995). Examination of pollen records in
conjunction with charcoal analyses provides information about vegetation response to fire.
Reconstruction of fire histories using charcoal
analysis is often combined with dendroecological
methods, in which comparisons of data from increment cores and cut sections of nearby fire-scarred
trees are used to date local fires (Clark, 1990). Both
methods have advantages and limitations. Fire-scar
279
chronologies are spatially more precise than charcoal analysis, but records become more sparse with
time and do not extend beyond tree life-spans, although records from snags and logs can lengthen
chronologies (Clark, 1990). Charcoal analyses can
provide temporal sequences of thousands of years,
although generally with less precise resolution than
fire-scar chronologies (Fall, 1997); they can also be
directly correlated with pollen records (Whitlock and
Millspaugh, 1996). Comparison of fire-scar dates
with charcoal profiles can be used to calibrate thresholds below which charcoal amounts are considered
background levels (Millspaugh and Whitlock, 1995).
Charcoal peaks above the threshold are interpreted
as fire events. Charcoal analysis can also provide an
important source of information about fire frequencies in systems in which trees are killed by fire and
thus do not provide fire scars for dating fire events
(Fall, 1997).
A number of studies have demonstrated a close
relationship between peaks in various charcoal particle size classes and occurrence of local fires (Whitlock and Millspaugh, 1996). However, no measure
of charcoal abundance has provided complete accuracy in detecting known local fires. No charcoal size
examined by MacDonald et al. (1991) consistently
identified all known local fires at Wood Buffalo National Park, Alberta, although the abundance of microscopic charcoal was found to be influenced by
regional fires. Charcoal deposition patterns are a
function of a number of factors, including the size,
intensity, fuel type, and meteorological conditions of
fires, as well as lake and watershed characteristics
(Patterson et al., 1987; MacDonald et al., 1991;
Whitlock and Millspaugh, 1996). In addition, charcoal studies have employed a variety of analytical
techniques. Given this variability in techniques, deposition patterns, and results, Whitlock and
Millspaugh (1996) recommend that different charcoal size classes and analytical techniques be compared in a fire history study to determine the best
method for detecting local fires in an area.
EXAMPLES
1. Millspaugh and Whitlock (1995) constructed
a 750-year fire history for the Central Plateau of
Yellowstone National Park. Regional and local
fires were distinguished by comparing the charcoal
record of a large lake with those of nearby small
lakes. In addition, local fires (within the watershed)
were discriminated from regional fires in charcoal
records by determining magnetic susceptibility,
which is high in burned soils eroded into lake sed-